Thanks to all of those who participated in our hardware failure survey. We’ve found some interesting data we’d like to share.
The goal of this survey was to determine which types of equipment fail most frequently, why they fail, and what the most difficult things about recovering these pieces of equipment are.
We all know that hardware failure is very common, so after polling nearly four hundred of our partners, it’s not surprising that 99 percent of them had experienced a hardware failure. On top of that, data loss is common for clients our partners work with, evidenced by the fact that the majority of IT professionals (71 percent) have worked with clients during a major data loss event.
So what causes failures? The sweeping majority (80.9 percent) of failures are caused by hard drive malfunction. If you look at the most common problems outside of hard drive failure, it’s a tossup. An indeterminate failure is nearly as likely to happen as a motherboard failure or other internal component failure (see chart “Most common hardware problems”). When it comes to failure in various types of server, it’s clear that no one type of server fails more than another, meaning all servers should be backed up—not just those that seem most critical to operations.
When we asked how much downtime clients our partners work with would accept (specifically domain servers), the plurality (40.8 percent) were comfortable with downtime in the two to four hour range, 14 percent said four to six hours was acceptable, 26.4 percent said one hour was acceptable, and another 14 percent said that no downtime was acceptable.
Users of StorageCraft ShadowProtect can easily recover within the two to four hour window, as illustrated by the fact that many respondents (40.6 percent) said they could recover a down server in under two hours, while a total of 67.2 percent said they could do it in under five.
We also asked respondents to tell us in their own words what the most difficult thing about recovering a failed machine was. A large number of respondents agreed that issues related to finding drivers and completing bare metal restores were the most difficult. The next most common answer was that issues related to transferring large amounts of data from hardware to hardware was the most difficult thing (usually because of the amount of time it can take for large amounts). Pressure from clients to hurry or unrealistic recovery expectations from clients was the third most common answer, suggesting that RTOs should be established at the onset, not during a recovery scenario, in order to manage client expectations.
Among StorageCraft partners,
– 99 percent said they had experienced a hardware failure in the past.
– The plurality (40.6 percent) can recover a downed server or desktop in under two hours using ShadowProtect
– 71.1 percent say they have worked with a client during a major data loss event.
– 14 percent said no domain server downtime was acceptable, while 40.8 percent said between two and four hours of downtime was acceptable.
– 52.7 percent said illustrating the value of disaster recovery to clients was the most difficult thing about a backup and disaster recovery offering, followed by the actual performance of failovers and recoveries (17.8 percent) and testing disaster recovery plans with clients (15.2 percent).
– Hard drive failure is by far the most common hardware problem (80.9 percent) followed by power source failure (4.7 percent).
These statistics are based on a ten-question survey of 387 StorageCraft partners. Research conducted by StorageCraft Technology Corporation.